HEDAS Vortex-Scale Data Assimilation with Aircraft and Satellite Retrieval Observations: Summary of the 2013 Atlantic Hurricane Season Results

The Hurricane Ensemble Data Assimilation System (HEDAS) was developed to assimilate high-resolution observations at the vortex scale. It combines a state-of-the-art square-root ensemble Kalman filter, National Oceanic and Atmospheric Administration's (NOAA) Hurricane Weather Research and Forecasting (HWRF) modeling system, and a storm-relative processing capability for a variety of observation types. In 2013, HEDAS was run in near real time on the NOAA Hurricane Forecast Improvement Project's Jet supercomputer to assimilate observations from NOAA and Air Force Reserve research and reconnaissance flights (dropwindsonde, flight level, Stepped-Frequency Microwave Radiometer, and Doppler radar data), and new observations from Global Hawk dropwindsondes, satellite Atmospheric Motion Vectors, and retrieved thermodynamic profiles from the Atmospheric InfraRed Sounder and Global Positioning System Radio Occultation. For the 2013 hurricane season, 54 such TC cases were run with various combinations of these observation types.

A summary of the 2013 hurricane season HEDAS runs will be presented. Both HEDAS final vortex analyses and deterministic HWRF forecasts initialized with these analyses will be discussed. A comparison will also be made to when HEDAS is run only with aircraft observations (satellite retrievals withheld from assimilation; standard implementation prior to 2013) so that the added impact of the satellite data can be assessed. Since aircraft reconnaissance flights have limited geographical range, platforms such as satellites and unmanned long-range aircraft allow for coverage of TCs far from land bases and hypothetically allow for continuous cycling at the vortex scale. With this motivation, cases for which only satellite data and/or NASA Global Hawk observations were available will also be investigated separately to assess the value of such platforms for vortex-scale data assimilation.